Cybersecurity Trends is an Annual hot topic amongst bloggers, news reporters, conference sessions, Ted Talks and a million other outlets. This year’s entry’s all mimic each other when listing the top movements:

• IoT and Cloud Security
• Work From Home Cybersecurity Challenges
• Cultivating a Security Aware Culture throughout the Organization
• State Sponsored Attacks are now targeting Businesses at an equal rate.
• Realtime Data monitoring and safeguards
• Targeted Ransomware will continue to rise significantly.

But I believe there is one trend, specifically in North America but also globally, that out paces all of these. Particularly since it weaves its way into all of them. And that’s the rapidly growing landscape of Artificial Intelligence (AI). But this is VERY much a double-edged sword. AI can quickly and efficiently crunch vast amounts of data points in order to predict, thwart, detect and respond to threats. Conversely, the bad actors of the world are also using AI to become more proficient at:

• Identifying weaknesses and ingress points
• Create incredibly clever and realistic Phishing attacks.
• Automating and coordinating their attacks.
• Creating Synthetics Identities to use for various nefarious reasons.
• Manipulate Social Engineering scripts.
• More efficient Password Hacking

Let’s talk about how AI can HELP your cybersecurity posture. The number of areas where AI can positively effect our industry’s capabilities against the hackers is equally lengthy and impressive.

• Automated Testing to look for areas of weakness.
• Ability to identify patterns and diagnose attacks in realtime
• Respond to incidents in realtime
• Behavior analysis of both humans and transactional data
• Advanced Authentication factors including Biometric Liveness Detection methods, face and voice biometric accuracy levels and behavioral biometrics

There are those that call this double-edged sword an “arms race”. So how do you stay ahead in an arms race when the machines are learning on their own to battle each other?? Well to answer that, you need a better understanding of how AI works. It’s not magic. An AI/ML engine needs to actually be “taught” things initially and continually above and beyond it’s own learning. These engines rely on algorithms and statistical models (advanced ones are call Neural Networks) to analyze all sorts of data and generate decisions based off that. Which means, they need to be taught by entering large datasets to train the system. Once trained sufficiently, an AI/ML engine can be deployed. But just like us humans, these engines need to continually advance and adapt to new data. And thus will only be as good as the datasets at their disposal. Otherwise an AI/ML engine can be accused of being just like me… using too much “old school thinking”.

So they key to winning an AI/ML arms race is to consistently increase the quantity AND quality of input data.  And quality equals diversity of datasets also! This is where I think we can zero in on a key piece of ammunition in the arms race. That quality and diversity of data that gets inputted into the AI/ML engines NEEDS to come from a concerted effort of collaboration from the cybersecurity community. Forensics investigations and the learning from them need to be disseminated and shared with these engines to make them better! And this needs to happen from a Global aspect. Patterns of attacks in Asia have a different flavor to them then the attacks in North America or Europe. The attacks into Government sites have a different approach then the ones to the financial industry or the supply chain. But ALL of these flavors need to make it into our AI/ML engines to help us as a community thwart the constant barrage of attacks on a daily basis. And that’s why I think that AI will be the very top Cybersecurity trend in 2023!

SISA is leading the way in this regard. Our company mission includes sharing our vast global knowledge leadership with the cybersecurity community. SISA is in a unique position based on the amount and variety of worldwide investigations that we perform. Not only do we feed that knowledge leadership back into our services, but we feel it’s a responsibility to disseminate in many different forms…like our Annual Top 5 Learnings report. It’s often said that humans are always the weakest link in cybersecurity. Well I (and SISA as a whole) feel like humans can and ARE the strongest line of defense! When we share and collaborate…WE WIN THE ARMS RACE!